Data from the Early Pregnancy Study (EPS) (A.Wilcox, PI) provide a resource for ongoing research into reproductive hormones, fertility, & early pregnancy. During field work for this study in 1982-1985, women enrolled at the time they stopped using birth control in order to conceive. We followed them through their 8th week of pregnancy. They collected daily first morning urine specimens and these were analyzed for human chorionic gonadotropin and steroid metabolites. Women with known fertility problems were excluded, so the sample represents normal unassisted reproduction. We conducted a pilot study to test the stability of hormones in urines stored from the Early Pregnancy Study. Pilot work supported the validity of BPA and phthalate measurements in the EPS urines after long-term storage, so we have designed a study to look at the association of those exposures with fertility and pregnancy outcomes in EPS. We designed a sample selection protocol (pool of 3 separate daily specimens) to the assess levels in each participant menstrual cycle and each clinical pregnancy. These have been analyzed at CDC. In preliminary analysis of the data, there is no evidence of any associations between BPA or phthalates and early pregnancy loss or time to pregnancy. Epidemiologic study of menstrual characteristics, fertility, and miscarriage relies on self-report because these outcomes are not systematically monitored by medical care protocols. Therefore, study validity depends upon the accuracy of self-report for these outcomes. We have evaluated the validity of self-reported cycle characteristics by comparing interview data to prospective daily recording of menstrual bleeding and pain. In another methodologic project, we are using available data to determine how well women can report the timing of their positive pregnancy test. Bias arises in spontaneous abortion & time-to-pregnancy studies when comparing exposed and unexposed groups that differ in timing of pregnancy testing. Those who tend to have delayed pregnancy testing will recognize fewer spontaneous abortions, and they will have longer times to pregnancy. In another analysis we are comparing self-reported retrospective data on time to pregnancy with prospectively-collected data (documented for participants in the Early Pregnancy Study during their participation). Participants were traced and sent a self-administered questionnaire to collect their retrospective self-reports. These data are currently being analyzed. We also studied length of human gestation using EPS data. In the EPS followup we collected detailed delivery data for the EPS pregnancy, so that we could identify which pregnancies ended with spontaneous labor and which ended as a result of medical intervention (induction of labor on C-section before labor). Such medical intervention artificially shortens gestation and must be taken into account when studying human gestation. We found remarkable variability in the natural length of gestation (spaning 5 weeks). Early pregnancy events such at timing of implantation and type of corpus luteum rescue were associated with length of gestation. I have done a comprehensive literature review of female fecundability studies (to be published as a book chapter in a new edition of Women & Health). We continue our research into factors affecting fertility. We are looking at pesticide exposure using data from the Agricultural Health Study. In addition, we have a special interest in female age-related decline in fertility. An ongoing prospective time-to-pregnancy study is providing data on antimullerian hormone as a potential marker of fecundability. The potential for this hormone to be used as an ancillary or primary outcome measure for studies of female fecundability needs further investigation, and the need for this new area of investigation has been presented in a commentary. We have examined the effects of early life exposures on age of menarche using data from the Sister Study. In work as a collaborator on Kathie Hartmanns Right From The Start study, we have found that early pregnancy miscarriage rates are very similar for African American and white women, but African Americans have nearly twice the risk of miscarriage during gestational weeks 10-20 compared to white women. Data from this study also demonstrated no evidence of early pregnancy non-steroidal anti-inflammatory medication being a risk factor for miscarriage.